Page Speed Insights


(Trey Yancy) #1

I think it woiuld be a good idea for everyone to check their speed ratings via the following link:

https://developers.google.com/speed/pagespeed/insights/

Some might be surprised at their ratings.

The issues with loading speed are knocking my ratings down for above-the-fold. For me, loading speed is the number one most critical issue with Hype. I don’t want to be the squeaky wheel, but the ability to run a Hype animation while simultaneously loading content is important and in many cases (if not most) this can not be remedied by a preloader or by splitting things into scenes and delaying loads through that method.

A related issue is that of reloading JS and duplicate content when you hit a new page. A shared site-wide resource folder would address the latter issue. For the former, some means of consolidating JS would be helpful (particularly when there are several Hype items on the same page). These two things are what are responsible for giving me a 49% speed score on various pages.

I want Google to like my sites, but I want the sites to be magnetic as well. Boxey, static, bootstrap-style sites are everywhere. Minimizing the Hype content seems to be the only way to achieve a good speed rating.

If this is seen to be an issue with a sizable number of Hype users, perhaps it could become a top priority for the development team.


(𝕄𝕚𝕔𝕙𝕒𝕖𝕝 𝔾𝕒𝕣𝕠𝕗𝕒𝕝𝕠) #2

I’ve suggested to Tumult that Hype do some lossless compression on the images in Hype.

Meanwhile, I use ImageOptim.

Hey @jonathan - looks like they have a server API now… https://imageoptim.com/api


(Jonathan Deutsch) #3

That’s pretty cool, though I don’t think we could use this general-purpose due to offline needs and not all users would want their data sent to a 3rd party! ImageOptim is a great project but it doesn’t look like they only have a GPL license.


(𝕄𝕚𝕔𝕙𝕒𝕖𝕝 𝔾𝕒𝕣𝕠𝕗𝕒𝕝𝕠) #4

No, it looks like another “cloud” solution. Although, maybe Hype could have its own image optimization.

SVGO - https://github.com/svg/svgo

Also, ImageOptim is on GitHub too… https://github.com/ImageOptim/ImageOptim


(Jonathan Deutsch) #5

Yeah - I’d like to include better image optimization as part of the export. While ImageOptim is on GitHub, their license is incompatible with usage in Hype. However, it does rely on some libraries like pngcrush that do have a compatible license. (Another option is always contacting the developers and seeing if they will grant a different license for me, but that is usually unlikely or expensive!).


(Trey Yancy) #6

TinyPNG / TinyJPG works very well. I set up a couple of actions in Photoshop that will zip through a ton of images quickly, put them in a holding folder and close the processed files. I could then put them where needed (for example, processing the images in a Hype resource folder, copying them back to all instances of the folder, as well as processing the source images). Last month I went through all my images created over the past six years (those that did not have a required color profile) and processed the lot. It was somewhere in the neighborhood of 5K images. Well worth the $25.

The tools are available. I would put image compression low on the list. An updated loading system would make a huge difference - particularly with Google speed ratings. The above the fold Hype content on my sites has knocked the daylights out of my speed ratings and, as a consequence, knocked me down in the search results. It is urgent for me because I want the cool stuff to be at the top of the page and not pushed down the page below mundane photos and text.

This aspect of the ranking issue, which is based on Google’s UE algorithms, is what makes this so house a’fire urgent. There are various tricks, such as delaying image loading by scene, having the background image on the page beneath a transparent Hype animation, etc. but that will only get you so far.

The iPhone has been out for 9 years now, with corresponds to the time that a serious Flash replacement has been a priority. The simple fact is that Tumult has to maintain a balance between its development budget and the business of staying in business. They can only do so much. It is none of my business to offer advice on how things are done, but I would say that practical matters, such as Google ratings and rankings - things that affect users’ bottom line - have a bit more weight than coolness. Coolness drives sales, so updates must incorporate these things, but enhanced loading to boost Google ratings, is also a compelling feature.


(Jonathan Deutsch) #7

I agree; a lot of the holdup has been on dissatisfaction on an good design for this. But this post got my brain working on it again to flesh out some more ideas (based on using the “load resource” action we chatted about a while ago), and I think I finally have a solution that is simple and offers all the features users have requested over the years in regards to loading/preloading :slight_smile:


(shaun) #8

I’m reading this and thinking 'And… Where is it, did you just let it slide again?
TYancy was right this is very important.
Did you get anywhere with it?


(Jonathan Deutsch) #9

Greater preloading is not yet a part of Hype. We try to work on items in a priority and efficiency based method and don’t let things slide :smile:.


(Trey Yancy) #10

After my years as a product reviewer (and compensated beta tester for various companies), I know that Tumult has several things in common with other developers of this kind, in that nothing slides and projects are handled based on a balance the priorities of consumer demand, feasibility, bang for buck (development cost vs rewards), staffing / scheduling, bug fixes and more (and not in this specific order). If money were no object, a developer could hire and train a ton of coders and blow their entire budget on staff.

Typically, for a company the size of Tumult, the biggest limitation when the rubber meets the road is marketing and sales. Back in the day before Internet dominated everything, a company could take out ads in MacUser, MacWorld and others, and know that the ads were being seen. These days, it is no longer that way. As for companies that do advertise in print, they have to pay exorbitant fees for ads that are viewed by a limited audience. Pay for click is a lot tougher, and not only because of the cost, but because the targeting must be so narrow (individual and corporate Mac users who are into motion graphics, video, and web) and because it is not easy for outfits such as Google to identify such a precise group (even if there were a Windows version). In addition, the (expensive) cash cow that was MacExpo is no more, so how do you put the product before the masses? Quite often it happens through partnering with companies that produce complementary products (such as the WYSIWYG web design apps), which also means offering a discount on the product.

So how does this relate to the topic under discussion? From my perspective, I am absolutely confident that the Tumult team is doing their very best to juggle all the above-mentioned concerns while working to increase the revenue that would make it possible to increase the level and range of feature development. And what can we, the users do about it? We can tell others about Hype and Tumult’s other products in order to drive revenue which, in turn, would drive development.


(Trey Yancy) #12

It all comes down to final project size. Optimizing only takes you so far. My projects tend to be commercial in nature, focused on building excitement for high-dollar products of a type that require a good percentage of large photos. The demographic is 85% female, aged 30 to 50, 90% of whom are strongly left-brain dominant. This means lots of beauty shots and limited text. I don’t think I represent the average Hype user, but there is some overlap with the wish lists.


(Ken Heins) #13

A very strong agreement on tinypng. Try it, you will be amazed at the amount of compression you can get. Sometimes 80% on a percentage basis, and if there is any loss in image quality, I am not seeing it.


(Trey Yancy) #14

Make sure to check the results. I have used this plug-in on a couple of thousand files and with many of them I found that the saturation was kicked up. I had to go through a bunch of these files and batch the saturation.


(Ken Heins) #15

I have probably done a hundred in the last month. I can’t see anything visually. How much of a numerical difference? I do a lot of galleries and most of my individual images are 72 dpi, seldom more than 300x400 pixels. Not a lot of things there to either improve or to harm.

Are you talking about a Iarger size and higher resolution initial image maybe?

I should tell you that I come from a screen printing background and if the tonal ranges are narrowed, it doesnt bother my “quality assurance” eyes, since in that world we could not actually simulate more than 8 separate color shades within each print color. And by nature, each of those 8 ranges would lean toward higher rather than lower saturation. So you might see a problem that I would not. It would be an actual problem for you where it would not have been for me. What I did was a cruder form of printing, narrower, more saturated ranges are necessary there.


(Trey Yancy) #16

I started out at the age of 15 building a press in my parents’ garage from scratch, cutting film by hand, and printing T-shirts to promote local businesses. By the time I was out of it, I was setting my own dot screens in Photoshop and such.

I didn’t have any issues with pngs, but the jpgs were kicked up quite a bit (tilted slightly in the direction of My Little Pony cartoons). I ended up creating a batch that would convert back to RGB, drop the saturation by a couple of points, convert to index and then output. Once I had the basic action set up, I just dragged the culprits into a folder and hit the go button. At least I didn’t have to slave over them.


(Ken Heins) #17

Oh wow, we discovered the same solutions, (I actually think I read about it in a screen print magazine.) Then when sublimation dye inks came out, we were able to create a white background on dark shirts, flash the white sub ink to get a white background and then printed 6 color spot plastisol process just as if the shirts were white. We rotated the index color outputs by 7 degrees which put the dot patterns out of phase with the angle of the mesh, it worked great. We had a totally paid for automatic that would turn out 400 shirts an hour. Probably should not have sold that business, but thats what divorce will do to you. And being totally burned out after 19 years!

What a cool thing that I would run into someone with the same background.


(Trey Yancy) #18

I started a magazine that focused on scholarly articles on a huge range of interesting subjects that were written for laypersons - aimed at the “renaissance mann / woman”. I had 300 copies printed of the first issue and gave them away for free, paying for the printing from the subscriptions. Within four months I had readers on four continents. This was with a zero budget for advertising. After two years of doing this on top of my day job (ninety-hour work weeks) I had just reached the point of quitting my day job and was looking for satellite printing houses in Australia and Asia. Then I was approached by a group who offered to buy the magazine. I was dealing with burn out. If I had hung on for another year, things would have become much more manageable, but I made the decision to sell the magazine. I later regretted it. The new owners didn’t have what it takes and they ceased publication shortly after taking over.


(𝕄𝕚𝕔𝕙𝕒𝕖𝕝 𝔾𝕒𝕣𝕠𝕗𝕒𝕝𝕠) #19

I too am battling with burnout. HA! Perhaps this could be its own thread. :smile:

I’m actually sitting on a pretty good idea. It takes a modest – but significant – investment of money, and an even greater investment of time and effort. I can do it – except I feel so tired all the time. HA HA.

Seeing your comments makes me wonder if I should keep fighting anyway.


(Ken Heins) #20

Yes, you have to keep fighting, if its a good idea, you will always wonder what would have happened.